However:  I do not know enough physics to know if it is possible; only enough to know how I would try to do it.  The basic idea would be to use the difference between the interference patterns that can be generated when particles do not choose, versus the lack of interference pattern that occurs when they are forced to choose.  Obviously the particles would be in batches--sequences--long enough to reliably create such patterns.  

Of technical problems, there should be many.  

Theoretically, it would be very interesting.  

The earth is only 21 light-milliseconds across.  I don't think it will improve cellphones much.  

The Fates are kind.

Any experimental verification of superluminal communication?

Kind of.

And also here.

I can understand the group velocity argument, but what bothers me is that the rationale for 'proving' that superluminal communication can't happen seems very similar to the one that was being used to 'prove' with absolute relativistic certainty that the group velocity has to be less than c.

This doesn't 'prove' anything about what's possible, but it does make me suspicious of the rigour of the arguments that are being used.